Clarify maximum sizes

[gustavdelius]

We currently have a confusing number of parameters related to maximum size:

• w_repro_max The size at which mature individuals invest 100% of their energy into reproduction. Note that there is not guarantee that all individuals are mature at this size. However we currently set the proportion of energy invested by all fish of that size and above to 100%, see What to do if not all fish are mature at maximum size? #300. So currently no fish should be able to grow beyond w_repro_max, but see Round down w_repro_max to next-lower point on w grid #301.
• w_max The size beyond which we do not believe the fish will ever grow in our scenarios. Used to choose the range of plots. Note that w_max could be smaller than w_repro_max for fish with high mortality, see Issue warning when a species abundance is not negligible above its w_max #298
• w_inf The von Bertalanffy parameter giving the asymptotic size of the von Bertalanffy growth curve, which is supposed to represent an average fish.

Note that none of these parameters set the actual maximum size of fish. That depends on growth and mortality rates and thus changes dynamically and needs to be read off from the actual abundances.

We currently have a complicated system of providing defaults for those maximum size parameters that the user does not specify explicitly.

When working on this also look at 1906de5